Thesis
Reference
Mesenchymal stromal cell interactions with pancreatic islets and liver cells modulate viability and functionality
MONTANARI, Elisa
Abstract
Multipotent mesenchymal stromal cells (MSC) seems promising for the treatment of numerous diseases. However, mechanisms leading to beneficial effects remain obscure. We explored the potential of human MSCs on the function of human islets and porcine hepatocyte viability.
Results show that culturing islets with MSCs promoted insulin secretion solely under cell-to-cell contact condition. The involvement of N-cadherin interaction in this beneficial effect was identified since its specific blocking abolished the MSC-induced insulin secretion.
Transplantation of co-encapsulated islets and MSCs into immunocompetent and streptozotocin-induced diabetic-mice significantly prolonged graft survival and function.
Further, a high-yield isolation protocol for porcine hepatocytes was established. Co-culturing and co-encapsulation of hepatocytes with MSCs improved hepatocyte survival and function, as demonstrated by the increased metabolic function and albumin secretion of hepatocytes in vitro. Our results suggest that beneficial effects by MSCs are mediated by N-cadherin interactions and that MSC are efficient to improve islet and [...]
MONTANARI, Elisa. Mesenchymal stromal cell interactions with pancreatic islets and liver cells modulate viability and functionality. Thèse de doctorat : Univ. Genève, 2017, no. Sc. 5102
DOI : 10.13097/archive-ouverte/unige:96379 URN : urn:nbn:ch:unige-963794
Available at:
http://archive-ouverte.unige.ch/unige:96379
Disclaimer: layout of this document may differ from the published version.
1 / 1
UNIVERSITÉ DE GENÈVE
Département de Génétique et Evolution FACULTÉ DES SCIENCES
Professeure Brigitte Galliot
Département de Chirurgie Viscérale FACULTÉ DE MÉDECINE
Professeur Leo H. Bühler
Mesenchymal stromal cell interactions with pancreatic islets and liver cells modulate viability and functionality
THÈSE
présentée à la Faculté des sciences de l’Université de Genève pour obtenir le grade de Docteur ès sciences, mention biologie
par
Elisa MONTANARI de
Reggio Emilia (Italie)
Thèse n° 5102
Genève Unicopy
2017
2 UNIVERSITÉ DE GENÈVE
Département de Génétique et Evolution FACULTÉ DES SCIENCES
Professeure Brigitte Galliot
Département de Chirurgie Viscérale FACULTÉ DE MÉDECINE
Professeur Leo H. Bühler
Mesenchymal stromal cell interactions with pancreatic islets and liver cells modulate viability and functionality
THÈSE
présentée à la Faculté des sciences de l’Université de Genève pour obtenir le grade de Docteur ès sciences, mention biologie
par
Elisa MONTANARI de
Reggio Emilia (Italie)
Thèse n° 5102
Genève Unicopy
2017
3
4
TABLE OF CONTENTS
1. Abbreviations ... 6
2. Summary ... 8
3. Résumé en français ... 10
4. Introduction ... 12
4.1 Multipotent mesenchymal stromal cells ... 12
4.1.1 Secreted molecules and paracrine effects ... 14
4.1.1.1 Effect on cell viability ... 15
4.1.1.2 Angiogenic effect ... 15
4.1.1.3 Anti-fibrotic effect ... 16
4.1.2 Differentiation capacity ... 16
4.1.3 Immunomodulatory effects ... 19
4.1.4 Therapeutic potential ... 23
4.1.4.1 Type 1 and 2 diabetes ... 24
4.1.4.2 Liver injury ... 25
4.1.4.3 Side effects ... 27
4.2 Mesenchymal stromal cells and pancreatic islets ... 29
4.2.1 MSCs improve islet viability and functionality ... 29
4.2.2 Effects on islet vascularization ... 31
4.2.3 Mechanism of islet-MSC cell-to-cell contact ... 32
4.3 Mesenchymal stromal cells and liver cells ... 33
4.3.1 MSCs sustain hepatocytes ... 33
4.3.1 MSCs improve hepatocyte viability and functionality ... 34
4.4 Cell encapsulation ... 36
5
4.4.1 Islet encapsulation and transplantation outcome ... 37
4.4.2 Hepatocyte encapsulation and transplantation outcome ... 38
5. Results ... 40
5.1 Aims ... 40
5.2 MSCs and human islets of Langerhans ... 42
5.3 MSCs and porcine hepatocytes ... 80
6. Discussion and Perspectives ... 112
6.1 MSCs and islets of Langerhans ... 113
6.2 MSCs and porcine hepatocytes ... 116
6.3 Perspectives ... 118
6.4 Conclusions ... 120
7. References ... 122
8. Appendix ... 140
8.1 “Microencapsulation of hepatocytes and mesenchymal stem cells for therapeutic applications” ... 140
8.2 List of publications ... 156
9. Acknowledgements ... 160
6
1. ABBREVIATIONS
ALT Alanine aminotransferase
AST Aspartate aminotransferase
CCL4 Carbon tetrachloride
CD Cluster of differentiation
CK Cytokeratin
DC Dendritic cells
E Epithelial
GVHD Graft versus host disease
HGF Hepatocyte growth factor
HLA Human leukocyte antigen
HNF Hepatocyte nuclear factor
HSC Hematopoietic stem cells
ICAM Intercellular adhesion molecule
IDO Indoleamine 2,3dioxygenase
IFN Interferon
Ig Immunoglobulin
IGF Insulin-like growth factor
IL Interleukin
ISCT International Society of Cellular Therapy
MCP Monocyte chemoattractant protein
MFGE8 Milk fat globule-EGF factor 8
MHC Major Histocompatibility Complex
MMP Matrix metalloproteinase
7
MSC Multipotent mesenchymal stromal cell
N Neural
NK Natural killer
NOD Non-obese diabetic
Pdl Programmed death-ligand
PDX-1 Pancreatic and duodenal homeobox 1
PEG Poly(ethylene glycol)
PGE2 Prostaglandin E2
ROCK Rho associated kinase
STZ Streptozotocin
T1D Type 1 diabetes
TGF Transforming growth factor
Th T helper
TIMP Tissue inhibitors of metalloproteinases
TNF Tumor necrosis factor
Treg T regulatory
VEGF Vascular endothelial growth factor
8
2. SUMMARY
Mesenchymal stromal cells (MSCs) actively contribute to their environment by secretion of cytokines, growth factors and extracellular matrix molecules. Their contribution to tissue regeneration remains to be explored. Likewise, a better understanding of molecular mechanisms leading to the beneficial effects in tissue repair processes will ultimately help to define the functional role of MSCs and their use for regenerative applications.
In this study, we first explored the potential of human MSCs to sustain the function of isolated human islets of Langerhans. Co-culture studies revealed that MSCs, together with islets, promoted insulin secretion. This effect was dependent on cell-to-cell contact, and was absent when islets were cultured alone or in co-culture with MSCs without direct contact.
Using qPCR analysis, we identified expression of the adhesion molecule N-cadherin in islets and MSCs. Specific blocking antibodies against N-cadherin significantly decreased the MSC- induced increase in insulin secretion, without affecting insulin secretion by islets alone.
Furthermore, the transplantation of co-encapsulated islets and MSCs into immunocompetent and streptozotocin-induced diabetic mice revealed a significant prolongation of graft survival, compared to mice transplanted with encapsulated islets alone.
Human MSCs were investigated for the improvement of porcine hepatocyte survival and albumin secretion. To this aim, we optimized and standardized a high-yield isolation protocol for hepatocytes from porcine liver. We demonstrated in vitro that MSCs enhance both the metabolic function of hepatocytes and albumin secretion. Co-encapsulation of hepatocytes with MSCs, using a novel biocompatible PEG hydrogel, showed increased viability and albumin secretory functions.
9
In conclusion, our data suggests that MSCs provide beneficial structural and paracrine signals for both human islets and porcine hepatocytes. As a mechanism, we reveal that MSCs increase islet function via N-cadherin interactions, demonstrating that paracrine signals are insufficient on their own. Therefore, co-encapsulation of hepatocytes and islets with MSCs might represent a valuable strategy to increase the viability of cells in microcapsules for clinical approaches, for future treatments in T1D and acute liver failure.
10
3. RÉSUMÉ EN FRANÇAIS
Les cellules multipotentes mésenchymateuses stromales (MSC) contribuent activement à leur environnement en sécrétant des cytokines, des facteurs de croissance et des molécules de la matrice extracellulaire. Toutefois, leur contribution à la régénération tissulaire est encore à explorer. Une meilleure compréhension du mécanisme moléculaire menant à des effets bénéfiques dans le processus de réparation des tissus permettrait ainsi de définir le rôle fonctionnel des MSC et leur utilisation pour des applications de régénération.
Dans cette étude, nous avons d’abord exploré le potentiel des MSC humaines pour soutenir les fonctions des îlots de Langerhans humains. Des études ont révélé que les MSC, en co- culture avec les îlots, améliorent la sécrétion d’insuline. Cet effet est absent quand les îlots sont cultivés seuls ou en co-culture sans contact direct, il est donc dépendant du contact intercellulaire. Par analyse de qPCR, nous avons identifié l’expression de la molécule d’adhésion N-cadherine dans les îlots et les MSC. Des anticorps bloquants spécifiques anti-N- cadhérine diminuent significativement l’augmentation de sécrétion d’insuline induite par les MSC sans affecter la sécrétion d’insuline des îlots seuls. De plus, la transplantation d’îlots et de MSC co-encapsulés, dans des souris immunocompétentes et rendues diabétiques par injection de streptozotocine, a montré une prolongation de la survie du greffon en comparaison avec des souris transplantées avec des îlots encapsulés seuls.
De plus, nous avons optimisé et standardisé un protocole à haut rendement pour l’isolement d’hépatocytes de porc. Les MSC humaines ont été étudiées pour améliorer la survie et la sécrétion d’albumine des hépatocytes porcins. In vitro, les MSC soutiennent la sécrétion et la fonction métabolique des hépatocytes. La co-encapsulation des hépatocytes avec des MSC utilisant un nouvel hydrogel PEG biocompatible a montré une augmentation de la viabilité et de la fonction de sécrétion d’albumine.
11
En conclusion, nos résultats suggèrent que les MSC fournissent un bénéfice structurel ainsi que des signaux paracrines aux îlots humains et aux hépatocytes porcins. Concernant le mécanisme, nous avons révélé que les MSC augmentent la fonction des îlots via des interactions utilisant la N-cadhérine, démontrant ainsi que les seuls signaux paracrines ne sont pas suffisants. Par conséquent, la co-encapsulation des hépatocytes et des îlots avec des MSC peut représenter une stratégie valable dans le cadre clinique pour augmenter la viabilité des cellules dans les microcapsules pour de futurs traitements du diabète de type 1 et de l’insuffisance hépatique aigue.
12
4. INTRODUCTION
4.1 Multipotent mesenchymal stromal cells
Mesenchymal stromal cells (MSCs) were discovered by Friedenstein in 1970, when he isolated adherent fibroblast-like cells from the bone marrow of a guinea-pig (1). MSCs originate from the mesoderm and are now isolated from almost all tissues throughout the body (2). MSCs reside in the stromal adherent fraction of the bone marrow, where they sustain the homeostatic turnover of non-hematopoietic stromal cells, thus regulating hematopoietic stem cells (HSCs) maintenance (2-4). In fact, after bone marrow irradiation in mice and consequent death of all progenitor cells, the infusion of MSCs together with HSCs improves the recovery of hematopoiesis, compared to mice infused with HSCs alone (5). Similarly, in vitro cultured MSCs sustain hematopoiesis for up to six months (6). HSCs are sustained by several cells derived from MSCs, such as reticular cells, adipocytes and osteoblasts, as well as macrophages (7). These cells together constitute the hematopoietic stem cell “niche”, which regulate HSC function through cellular contact and paracrine signaling (8). Indeed, MSCs secrete many cytokines that promote cell renewal, such as stem cell factor (SCF), leukemia inhibitory factor (LIF), stromal cell-derived factor-1 (SDF-1), transforming growth factor (TGF)-β and others that promote hematopoietic cell maturation, such as granulocyte- macrophage (GM) colony-stimulating factor (CSF) and G-CSF (9). Hence, MSCs sustain HSC function by regulating quiescence and self-renewal through cell contact and secreted molecules (10).
MSCs are preferentially isolated from the bone marrow, where they represent 0.01%-0.0001%
of the nucleated cells (11). Given that MSCs are scarce in the bone marrow, MSCs are often isolated from several other abundant tissues, most commonly the adipose tissue (12-16), the umbilical cord (17, 18) and Wharton’s jelly (19), but also from the amniotic membrane (20),
13
placenta (21), dental pulp (22), tonsils (15, 23), lung (24), pancreas (25), liver (26), dermis (27, 28) and skeletal muscle (29).
To have a consensus about how to identify MSCs, the International Society of Cellular Therapy (ISCT) defined cells as Multipotent Mesenchymal Stromal Cells when they express or lack combination of surface markers, listed in Table 1 (11, 30-34).
Positive
(ISCT) CD73, CD90, CD105
Other positive markers
CD13, CD29, CD44, CD46, CD51, CD54, CD55, CD59, CD106, CD146 (MCAM), CD166, CD271, ICAM-1, ITGA11, STRO-1, VCAM-1, CXCR4, Sca-1, Nestin, PDGF Rα, PDGF Rβ, Integrin
α1 Negative
(ISCT)
CD14-, CD34-, CD45-, CD79a-, human leukocyte antigen (HLA) class 2
Other negative markers
CD-11b-, CD19-, CD31-, CD36-, CD40-, CD80-, CD86-, von Willebrand factor-
Table 1. Surface markers of MSC.
MSCs adhere to plastic support under in vitro culture conditions. Indeed, MSCs are adherent cells, with proliferative and self-renewal capacities and a spindle shaped morphology similar to fibroblast, although they maintain limited cell sprouts compared with fibroblasts (31, 35).
Furthermore, the ISCT defined MSCs by their ex vivo capacity to differentiate into osteocytes, chondrocytes and adipocytes. MSCs have limited differentiation capacities compared to embryonic stem cells or with induced pluripotent stem cells (IPS) that have differentiation potential similar to embryonic stem cells. However, these extended differentiation capacities implicate a high risk of teratoma development which remains minimal in MSCs.
The tissue source of MSCs may direct their ability to differentiate. In fact, after isolation from the pancreas, one study showed that MSCs expressed embryonic markers, such as Oct-4
14 (Pou5f1-POU domain, class 5, transcription factor 1), Sox2 (sex determining region Y-box 2) and Rex-1 (zinc finger protein 42), and presented high proliferative capacities and telomerase activity, demonstrating their undifferentiated state (25). Independently, another group retrieved pancreatic MSCs, which were positive for classical markers, such as cluster of differentiation (CD)13, CD29, CD44, CD73, CD90, CD105, nestin and vimentin and negative for von Willebrand factor, CD31, CD34, CD45, cytokeratin (CK) 19 and CA19.9, suggesting that the origin of these cells was not ductal, endothelial or hematopoietic (14, 36).
Nevertheless, pancreatic MSCs expressed constitutively specific markers of differentiation (25) and had the potential to differentiate into adipocytes and osteocytes (14, 36). Moreover, after culture with specific induction factors, MSCs derived from the pancreas, expressed Pdx- 1 (pancreatic and duodenal homeobox 1), insulin, connecting peptide or C peptide and Glut-2 (glucose transporter-2); whereas MSCs derived from the bone marrow under the same conditions only expressed Glut-2 and insulin (37). These data suggest that the differentiation potential of MSCs is dependent on their origin. Nonetheless, MSCs derived from the pancreas or the bone marrow manifest the same immunomodulatory capacity to suppress T cells after anti-CD3 and anti-CD28 stimulation (38).
4.1.1 Secreted molecules and paracrine effects
The precise molecular mechanisms leading to the beneficial effects by MSCs remain largely unknown. However, previous studies have identified some relevant molecules synthetized and released by MSCs. Indeed, MSCs secrete trophic molecules that could be classified into cytokines, growth factors, receptors and binding proteins (39). The principal trophic factors that are secreted by MSCs are: interleukin (IL)-6, TGF-β, hepatocyte growth factor (HGF), prostaglandin E2 (PGE2) and vascular endothelial growth factor (VEGF). Further, MSC-
15
conditioned medium or MSC-derived vesicles are responsible for proliferative, angiogenetic, anti-fibrotic and immunomodulatory effects (40).
4.1.1.1 Effect on cell viability
MSCs secrete trophic molecules that have the capacity to rescue injured cells, accelerate tissue repair and decrease apoptosis (40). A protein screen on a secretome derived from human MSCs, revealed a correlation between MSCs-derived VEGF and cell proliferation and development. Indeed, systemic injection of MSC-conditioned medium, led to increased survival in mice with acute liver failure (41). Moreover, VEGF has an anti-apoptotic effect (42). Additionally, HGF protects cells from apoptosis, while its neutralization suppresses the protective effects (43). HGF, together with VEGF and insulin-like growth factor (IGF)-1, protect renal cells from acute injury after ischemia reperfusion (44).
Furthermore, MSC-conditioned medium has the capacity to improve hepatocyte proliferation, stimulate angiogenesis, trigger anti-inflammatory cytokines and increase expression of hepatic genes relevant for proliferation, such as tumor necrosis factor (TNF)-α and IL-6 (45, 46). Moreover, MSC-conditioned medium increases in turn, VEGF and matrix metalloproteinase (MMP)-9 expression in hepatocytes, suggesting multiple mechanisms behind the increased cell survival and proliferation (45, 46).
4.1.1.2 Angiogenic effect
MSCs facilitate wound healing and are also implicated in promoting angiogenesis (39).
Several factors released by MSCs are known to promote angiogenesis. For example, VEGF, which increases endothelial cell migration and proliferation, thus promoting angiogenesis (39) as demonstrated in several animal models, such as dogs, rats and mice (42). Further, MMP-1 is involved in angiogenesis, in particular MMPs mediate the destruction of the basement
16 membrane, allowing endothelial cells to migrate and restore the capillary structure. Moreover, MMP-2 and MMP-9 derived from MSCs enhance elastin production during wound healing (47). Likewise, HFG is involved in the acceleration of the wound healing and angiogenesis (42).
4.1.1.3 Anti-fibrotic effect
Several studies report that MSC-conditioned medium exerts protective effects towards liver fibrosis in mouse animal models. Indeed, MSC-conditioned medium is efficient in decreasing α-smooth muscle actin, type I collagen and MMP-2 expression, molecules involved in fibroblast activation (31). Moreover, the secretome of umbilical cord-derived MSCs contains several proteins that are protective towards liver fibrosis in mice. In particular, milk fat globule-EGF factor 8 (MFGE8), injected in mice with liver injury, elicited anti-fibrotic effects that are reversed by the neutralization of the protein with a specific antibody. The hepatic level of MFGE8 in patients with cirrhosis was also reduced, correlating with the anti-fibrotic properties of MFGE8 (48). Further, IL-1Ra and stanniocalcin-1, derived from MSCs, are, in part, responsible for the reduction of fibrosis in bleomycin-induced lung fibrosis in mice (49, 50).
4.1.2 Differentiation capacity
MSCs are characterized by the capacity to differentiate into several cell types (figure 1).
MSCs originate from the mesodermal embryonic lineage, and have the capacity to differentiate principally into cells derived from the mesodermal lineage, such as osteocytes, chondrocytes and adipocytes, as well as muscle cells, myofibroblasts, cardiomyocytes and endothelial cells (31, 51, 52).
17
MSCs also have the capacity to trans-differentiate into cells derived from other germline lineages, as endoderm. MSCs can differentiate into islets (37) and hepatocytes (53, 54).
Furthermore, they have the capacity to trans-differentiate into cells derived from the ectoderm cell lineage, such as neurons and epithelial cells (11, 55). However, the crossing of the lineage barrier into cell types others than the mesodermal lineage remains controversial.
After in vitro differentiation of MSCs into osteocytes, calcium deposits are visible in the culture (31). The activation of the pathway leading to bone formation is well known. Runx2 (runt-related transcription factor 2), Osterix and β-catenin are the principal transcription factors involved in the osteogenic differentiation (56). Runx2 promotes MSCs to a pre- osteoblastic commitment, thus inhibiting chondrogenic and adipogenic differentiation (57).
Wnt, Notch and BMP activate Runx2, which also takes place through the translocation of Smad into the nucleus (58). Runx2 is essential for osteoblast differentiation since the suppression of Runx2 blocks Osterix, and its absence, leads in turn to arrest cortical bone and trabeculae formation in bones (59). β-catenin is regulated by Wnt and its inhibition prone MSCs to differentiate into adipocytes or chondrocytes (60). Additionally, CBF-1α (core binding factor-1 α) is involved in osteogenic differentiation (61).
Sox9 and ZNF145 (zinc-finger protein 145) are the principal transcription factors involved in chondrogenic differentiation (56, 62). Inhibition of ZNF145 diminishes chondrogenic differentiation, while its activation overexpresses Sox9 and chondrocyte formation (63).
Additionally, upregulation of FOXO3A (forkhead box O3) correlates with MSC differentiation into chondrocytes (64).
Accumulation of lipid droplets reveals MSC differentiation into adipocytes (31, 35). PPARγ-1 (peroxisome proliferator-activated receptor-γ), PPARγ-2 and EBF-1 (early B cell factor) are the principal transcription factors involved in the differentiation of MSCs in adipocytes (65-
18 68). Also, Twist-1, Sox2 and Oct4 expression is involved in adipogenic differentiation (51);
however, GATA2 and Foxa1 expression inhibits differentiation of MSC into adipocytes (69).
Several other transcription factors have been shown to be implicated in the differentiation of MSCs into cell types such as cardiomyocytes and skeletal muscle cells. Cardiomyocytes differentiate from MSCs via GATA-4 (70) and Nkx2.5 (Nk2 homeobox 5) (71, 72). GATA-4 is regulated by Wnt11 (73), and by histone acetylation, methylation and DNA methylation (74).
For differentiation into skeletal muscle cells, Pax3 (paired box protein 3), Pax7 (75), MyoD (myogenic differentiation 1) and Myf-5 (myogenic factor 5) are implicated (76).
Smooth muscle cells are differentiated via GATA6 and SRF (serum response factor) transcription factors (77).
Further, it has been shown that MSCs differentiate into cells normally derived from endoderm such as pancreatic islet cells and hepatocytes. MSCs derived from adipose tissue, human bone marrow, and tonsils can differentiate into pancreatic islet-like cells (12-15). It has been suggested that, rat bone marrow-derived MSCs can differentiate into insulin-producing β cells (78). Pdx1 positive MSCs are inclined to differentiate into functional insulin-producing cells, and Pax4 promotes Pdx1, the expression of which has been suggested to be important for differentiation into beta cells (79). Moreover, human adipose tissue derived MSCs can differentiate into insulin-, glucagon- and somatostatin-expressing cells (14).
Early studies showed that bone marrow derived MSCs also differentiate into hepatocyte-like cells after culture in Matrigel and culture medium supplemented with fibroblast growth factor-4 and HGF. Hepatocyte nuclear factor (HNF)-3β, GATA4, α-fetoprotein, transthyretin and, weakly, CK19 were expressed after 1 one week, whereas bi-nucleated cells expressing
19
CK18, HNF-1α and HNF-4 appeared after a few weeks. Cells secreted urea and presented cytochrome P450, suggesting a functional maturation of hepatocytes. In accordance to this, hepatocyte-like cells derived from rat and mouse MSCs secreted albumin, while human derived hepatocyte-like cells did not (54, 80).
Figure 1. Differentiation capacity of MSCs
Whereas the differentiation capacity of MSCs into osteocytes, chondrocytes and adipocytes has clearly been demonstrated, differentiation of MSCs into pancreatic beta cells, hepatocytes, cardiomyocytes, neurons and smooth muscle cells remains controversial and needs confirmation. Actually, most studies analyze gene expression after cell differentiation;
however studies demonstrating mature and fully differentiated cells are still missing.
4.1.3 Immunomodulatory effects
MSCs carry low immunogenicity, due to their low levels of human leukocyte antigen (HLA) class 1 and class 2 molecules expression. Moreover, they do not express co-stimulatory molecules that activate the immune system, such as CD40, CD80 and CD86 (81).
Furthermore, MSCs possess immunomodulatory functions towards the adaptive and the innate
20 immune systems, which occur via paracrine signaling or through cell-to-cell contact directly with immune cells (81-83).
MSCs have the capacity to suppress T cell activation and proliferation, for both CD4 and CD8 T lymphocytes (84) and this suppression is principally HLA independent (85). TGF-β, IL-10, indoleamine 2,3-dioxygenase (IDO) and HGF are involved in T cell inhibition. Indeed, their blocking in the MSC-conditioned medium, reversed suppression of T cell activation (81, 86).
T cell inhibition is also caused by monocyte chemoattractant protein (MCP)-1, secreted by MSCs (87), as demonstrated in encephalomyelitis mouse model (81). MSCs convert T helper (Th) 1 cells to a Th2 phenotype, a cell involved in an anti-inflammatory response, and further induce IL-4 secretion, a typical cytokine produced by Th2 cells (88). PGE2 and MSC-derived vesicles are involved in the switch to a Th2 response (81, 89). MSCs also inhibit Th17 functionality (90), but sustain function and proliferation of T regulatory (Treg) cells that immunomodulate the immune system (91, 92). The expression of T reg in vitro depended on cell-to-cell contact with MSCs and blocking of programmed death-ligand (Pdl)-1 impaired Treg proliferation, proving that Pdl-1 is involved in the supportive effect of MSCs on Treg cells (93, 94). MSCs exert the same effect in vivo, as shown in mouse transplant recipients treated with MSCs, where the proliferation of effector T cells was inhibited and Treg expansion is induced in the spleen, mesenteric lymph nodes and peritoneal lavage (91). Treg proliferation and function in vivo is mediated by IDO (95). T cells express CD25, which is the receptor of IL-2 that is a strong T cell activator (96). CD25 is blocked by MMP-2 and MMP- 9, secreted by MSC. Blockade of CD25 decreased T cell responsiveness and function (91, 97). Inhibition of MMP in vitro reactivated T cells (97, 98). The injection of an inhibitor of MMP-2 and MMP-9 was associated to earlier graft rejection in vivo, compared to non-treated mice (97). Hence, the neutralization of MMP-2 and MMP-9 inhibited the immunosuppressive effects of MSCs. Furthermore, MSCs can suppress the activation of CD8 cytotoxic T cell
21
(99); however, when CD8 T cells are activated prior to the incubation with MSCs, MSC- mediated-inhibitory effect is absent (100). Also, MSCs promote the differentiation of CD8 regulatory T cell that are known to inhibit lymphocyte proliferation (101).
MSCs suppress also B cell lymphocyte function, as demonstrated by the co-culture of MSCs and splenic B cells. After stimulation with lipopolysaccharide in vitro, the presence of MSCs suppressed B cell proliferation, as demonstrated by decreased presence of CD138 positive plasma cells, and immunoglobulin (Ig)M release. Furthermore, MSC-conditioned medium also inhibited differentiation and IgM and IgG secretion in B cells activated by T cell- dependent or T cell-independent antigen stimulation. Similar effects were described in mice, where the injection of MSCs decreased the accumulation of alloreactive antibodies (82).
These studies demonstrate that MSCs suppress B cells by trophic molecules in vivo; however MCP-1, IL-10, TGF-β, IDO seemed to play no role (102). Conversely, other studies report that IL-6, programmed death 1 pathway (103), IDO (104) and MSC-derived vesicles (81) are involved in B cell suppression. Nevertheless, there are some contrasting data reporting that MSCs support B cell proliferation and differentiation (105).
Natural killer (NK) T cells have a robust cytotoxic potential. MSCs modulate NK cytotoxicity in freshly cultured cells, decreasing proliferation and cytokine secretion, whereas in mature NK cells, MSCs exhibit no effect (104, 106). MSCs also modulate cytokine production and proliferation of NK cells (83).
In co-culture MSCs inhibit the capacity of dendritic cells (DC) to differentiate into mature and active DC. MSCs inhibit major histocompatibility complex (MHC) class II and CD1-α, which are molecules implicated in the antigen presentation, as well as co-stimulatory molecules CD40, CD80, and CD86 (107). MSCs can direct DC toward a tolerogenic phenotype, expressing decreased levels of MHC class II, with a diminished capacity to present antigens.
22 This inhibited CD80 expression and induced the PGE2, involved in the increased secretion of IL-10 in the conversion of DC into tolerogenic DC, and this has the potential to inhibit T and B cell activation in in vitro culture, and allogeneic islet transplantation in vivo (33). Moreover, MSCs diminish migration, endocytosis and maturation of DC (108, 109). MSC-derived vesicles co-cultured with DC, address the latter to an immature phenotype, with the reduction of activation markers and increased IL-10 and IL-6 production. This switch in DC marker expression and cytokine production drives DC to a tolerogenic phenotype, and these tolerogenic cells have the capacity to decrease Th17 and increase Treg cell numbers in in vitro co-cultures (110).
MSCs promote the differentiation of macrophages into M2 phenotype, thus promoting migratory capacities of monocytes and macrophages, and improving the phagocytic activity by cell-to-cell contact or by paracrine factors (83, 111). PGE2 is involved in macrophage regulation (112). MSCs decrease the secretion of inflammatory cytokines by activated macrophages (109); however MSCs boost macrophage activity, facilitating healing of wounds and injuries (113).
MSCs sustain neutrophil recruitment, by secreting cytokines that attract neutrophils and improve their functionality. STAT3 pathway is involved in the protection of neutrophils from apoptosis (114). Moreover, MSCs secrete IL-8 that modulate CD11b expression in neutrophils, regulating extravasation (115). MSCs improve neutrophil function in pathogen recognition; hence facilitating pathology resolution (83).
23
Figure 2. Immunomodulatory effects of MSCs on the different immune cells
4.1.4 Therapeutic potential
Autologous MSC transplantation could represent a solution for several pathologies, since MSCs exhibit limited immunogenicity, as they do not express HLA class 2, nor costimulatory molecules. For this reason, transplantation of allogeneic MSCs could represent an attractive therapeutic solution (4). Furthermore, MSCs have the potential to migrate towards injured sites and to alleviate the damage by proliferative, renewal, anti-apoptotic, anti-fibrotic and angiogenic capacities. These migratory capacities direct MSCs not only to the principal injured sites, but also to secondary injures, and allow the treatment of diseases that involve many organs (116).
MSCs have been successfully used for the treatment of osteogenesis imperfecta in children, where allogeneic MSC transplantation alleviated the symptoms of the pathology (117).
Further, MSCs have been efficient in several pathologies implying cellular regeneration, such as ischemic heart disease, cerebral and myocardial infarction (118, 119), stroke (120), spinal and cord injury (121), brain injury (122), corneal and retinal diseases, Crohn’s disease,
24 cartilage regeneration, acute and chronic liver failure (31) and ischemia reperfusion injury in the kidney and in the liver. MSCs were also effective in neurological and endocrinological pathologies, such as diabetes type 1 and 2 (17, 123-125). Moreover, since MSCs have immunomodulatory capacities, they have been applied for the treatment of autoimmune disorders such as lupus erythematosus systemicum (126-128) or of graft versus host disease (GVHD) (129).
GVHD is the immune reaction occurring after the transplantation of hematopoietic stem cells.
T cells from the graft react towards the antigen presenting cell of the host, leading to damage in several organs (130). Since MSCs elicit clear immunomodulatory capacities in vitro, several clinical studies are ongoing to treat GVHD; however, the beneficial effects remain controversial, despite the absence of adverse symptoms (129, 131 , 132-134). Mainly, the immunosuppressive effect seems to be principally realized by paracrine factors, such as TGF- β, IL-10, PGE2, IDO and HGF (135, 136).
4.1.4.1 Type 1 and 2 diabetes
Type 1 (T1D) and type 2 (T2D) diabetes are endocrine diseases, characterized by glycemia dysregulation, caused by autoimmune beta cell destruction (T1D) or insulin resistance associated to impaired insulin and glucagon levels, which are the hormones responsible for the maintenance of blood glucose homeostasis (137, 138).
Several studies observed beneficial effects of MSCs in humans. Intravenous infusion of umbilical-cord derived MSCs, administered twice at two-week intervals for 24 months, improved C-peptide and HbA1c1 levels in type 2 diabetes patients. Also, MSC transplantation reduced exogenous insulin supply without exerting any side effects (17, 139).
25
MSCs are also effective in the treatment of T1D and its complications. Indeed, using non- obese diabetic (NOD) mice, a mouse model of T1D, intravenous injection of undifferentiated MSCs, decreased glycemia and their survival was improved (140). Further, MSCs were detected in the pancreas after 23 days, where they differentiated into insulin-producing cells, as demonstrated by colocalization of GFP (fluorescent label marker of donor cells) and C peptide in the pancreas. MSCs reduced levels of reactive T cells, in particular Th1 and Th17, and increased levels of Treg in the spleen and locally in the pancreatic lymph nodes (19, 141).
These results suggest that MSCs have the potential to migrate to the damaged site, to differentiate into insulin-producing cells, to repair damaged cells and to also decrease the autoimmune reaction towards islets in T1D mouse model. Similar beneficial effects had also been reported in streptozotocin (STZ)-induced and high-fat-diet diabetic rats (142).
Interestingly, a study reported that the addition of liraglutide, an analogue of glucagon-like peptide-1, prolongs the ability of MSCs to preserve β cells in NOD mice (141).
It is also known that 20-40% of diabetic patients suffer of renal failure, as a consequence of diabetes (143). A study reports that MSC infusion in STZ induced-diabetic mice decreased levels of albuminuria and presented slight tubular dilatation in the kidney, suggesting that MSCs, after systemic injection, could migrate in the damaged sites and improve different injuries caused by the same pathology (144). A clinical phase 2 study revealed that infusion of MSCs in five patients with T1D and ketoacidosis significantly improved exogenous insulin uptake, with one patient becoming normoglycemic. Importantly, no patients developed side effects derived from MSCs (145).
4.1.4.2 Liver injury
Liver injuries comprise several pathologies that provoke damage to hepatic cell function.
Despite the regenerative capacity of the liver, some injuries result in an extensive hepatocyte
26 loss, which does not allow a sufficient regeneration of functional hepatocytes for the survival of the patient (146). MSCs have been investigated for their use as cellular therapy to favor regeneration and also to decrease liver fibrosis (147).
Hepatitis C virus could be responsible for liver damage, through the progressive transformation of liver parenchyma into fibrotic tissue (148). Autologous transplantation of MSCs in the parenchymal liver tissue of patients with cirrhotic liver, decreased hepatic enzymes, such as alanine aminotransferase (ALT) and aspartate aminotransferase (AST), bilirubin levels and extracellular matrix protein, leading to an improvement of liver function (149). MSCs can also reduce the hepatic fibrosis induced by the parasite Schistosoma mansoni (150). Furthermore, MSCs promote hepatocyte regeneration in a murine model of hepatic steatosis induced by a high-fat diet (151), however, a study report that MSCs exert no effects after infusion in the ischemia/reperfusion injury model (152). In cases of alcoholic cirrhosis, infusion of autologous MSCs improved the histological features, decreasing collagen deposition, α smooth muscle actin and TGF levels in 6 of 11 patients (153).
Acute liver failure consists in the fulminant destruction of hepatic cell and cause high mortality (154). MSCs have been infused in several animal models of acute liver failure, including mouse (155), rat (156) and pig (157) and MSCs exert anti-fibrotic and anti- apoptotic beneficial effects in all animal models used (158, 159). Recently, MSCs and MSC- conditioned medium were tested in a mouse model of fulminant hepatic failure and chronic liver fibrosis. Both were efficient in the recovery of hepatic failure; however MSCs were more efficient than MSC-conditioned medium in fulminant hepatic failure due to their immunosuppressive effects. On the other hand, MSC-conditioned medium was more efficient in chronic liver fibrosis since it decreased inflammatory responses (160). Contrarily to this, one study reported that MSC transplantation did not improve liver failure; however, systemic injection of extracellular vesicles from bone marrow-derived MSCs, reduced hepatic injury
27
and improved mice survival (161), demonstrating that microvesicles efficiently replace the paracrine effects of entire cells and suggesting that factors are transported and released via microvesicles. In light of these experiments, showing that paracrine effect seems to be more effective in rescuing liver failure, MSCs were encapsulated, prior to transplantation in mice with acute liver failure caused by bile duct ligation or carbon tetrachloride (CCL4)injection.
After intraperitoneal transplantation of encapsulated MSCs, collagen deposition levels in the liver were decreased, probably through an action on MMP-9 (31). These data strengthened the idea that the beneficial effects of MSCs on liver fibrosis occurs primarily through paracrine effects and do not necessarily need cell-to-cell interactions.
4.1.4.3 Side effects
Despite the beneficial effects that MSCs exert in several pathologies in experimental animal models, the risks of a cell therapy with MSCs needs to be evaluated. MSCs may present some risks related to malignant transformation, donor morbidity after MSC withdrawal, in cases of autologous transplantation and viral exposure (39).
MSCs are mainly isolated from tissues and need an extensive in vitro expansion, prior to therapeutic applications. High proliferation might cause cellular modifications. In the mouse model, MSC in vitro expansion led to malignant modifications (162, 163); however, human MSCs did not differentiate into malignant cells (164, 165). Moreover, recent research showed that now no malignant transformations had been described in clinical applications.
In case of auto-transplantation, MSCs are isolated from bone marrow aspiration via the iliac crest. These surgical procedures may cause complications or undesirable effects in patients, especially in immunosuppressed patients.
28 Contrary to auto-transplantation, allo-transplantation of MSCs involves the risk of the transmission of pathogens such as viruses. Studies report that after in vitro culture, MSCs were carrying cytomegalovirus, herpes simplex virus, parvovirus B19 and varicella zoster virus infections (166-168).
29
4.2 Mesenchymal stromal cells and pancreatic islets
Islets of Langerhans fulfill the endocrine function of the pancreas and are principally composed of insulin-secreting β cells and glucagon-secreting α cells. Insulin and glucagon are the key players for the maintenance of glucose homeostasis in the blood, the dysregulation of which causes diabetes (169). Morphologically, human islets are organized in substructures, where α cells line the endothelial cells and are surrounded by adjacent β cells (170). Islets are located in the exocrine tissue of the pancreas and are surrounded by a thin stromal layer and are highly vascularized. Cellular interactions between the different cell types include cadherins and integrins and are essential for the regulated insulin and glucagon release (171, 172).
4.2.1 MSCs improve islet viability and functionality
It is mainly established that the beneficial effects of MSC are mostly derived from the secretion of immunomodulatory and cyto-protective factors. However, in relation to islets, MSCs may also play a role in the conditioning of the micro-environment, especially in isolated islets which suffer from isolation with enzymatic digestion and from the absence of their natural micro-environment. Therefore, MSCs may contribute to an increased function by mimicking the natural environment of islets. Indeed, co-culture of islets and MSC, showed that cell dispersion and the sprout formation of islets were both improved (173). Further, insulin secretion by islets was strongly enhanced by MSCs, as demonstrated by co-culture models of MSCs and islets of mouse (174), rat (175, 176) and human (177). Cell-to-cell contact between islets and MSCs seems to be compulsory to improve insulin secretion and that culture of MSCs and islets without contact maintains insulin secretion similar to levels secreted by the islet alone (174). Co-culture of rat islets and MSCs showed decreased levels
30 of soluble MCP-1, TNF-α, and increased levels of tissue inhibitors of metalloproteinases (TIMP)-1 and VEGF after one month, however these results were not observed in transwell culture conditions, where there is only a paracrine exchange (175, 176). Further, the co- culture of rat islets and MSCs improved islet viability and increased the expression of insulin 1, Pdx1, platelet endothelial-cell adhesion molecule 1 and VEGFa, suggesting that the presence of MSCs facilitates insulin secretion and that the expression of growth factors decreases inflammation (178). However, there are some reports showing that mouse islets present enhanced survival only after exposure to trophic molecules secreted by human MSCs and not in cell-to-cell contact culture conditions. IL-6, IL-8, VEGF, HGF and TGF-β were detected in the MSCs conditioned medium and were suggested to be involved in the beneficial effects observed (179).
Moreover, MSCs preserved the function and viability of isolated islets injured by cytokines, such as interferon (IFN)-γ, TNF-α and IL-1β (177) and of islets cultured under hypoxia/reoxygenation conditions (180). Supernatants obtained from co-culture of MSC with STZ damaged rat islets showed increased levels of IL-6 and TGF-β. Furthermore, the expression of anti-apoptotic genes in the damaged islets positively correlated with the presence of MSCs, suggesting a role of MSCs in regulating the expression of anti-apoptotic genes (181).
MSCs co-transplanted in vivo with islets increase graft function and survival in syngeneic and allogeneic mouse and rat models (182-189). Also, islets pre-cultured with MSCs improved their outcome as a graft, demonstrating that exposure to paracrine factors exert potentiating effects on islets (174, 190). MSCs promote insulin secretion even in minimal mass islet transplantation studies. In such models, islets alone are incapable of reversing diabetes;
however, the presence of MSCs along with islets allowed the transplanted diabetic rodent recipients to reach normoglycemia (191, 192).
31
Some studies show that MSCs promote islet graft function even in large animal models, such as non-human primates, where the presence of MSC enhanced islet graft survival and decreased the number of islets necessary to reverse diabetes in allo-transplantation models (193). Similarly, in a model of human bone marrow MSCs co-transplanted with neonatal porcine islets in mice, graft functionality is improved and diabetes is reversed faster compared to mice transplanted with islets alone (194). MSCs co-transplanted with islets exert immunosuppressive effects, and the presence of MSCs decrease T cell recruitment around the graft, suppress Th1 cell activation and increase the percentage of Treg in the blood of transplanted animals (182, 193, 195, 196). Further, MSCs decreased the levels of follicular B helper T cells and auto-antibodies in the spleen and in the lymph nodes of recipients (196).
4.2.2 Effects on islet vascularization
Several studies in vivo suggest that MSCs increase vascularization in islets (182, 197, 198).
Also, MSCs differentiated into vascular endothelial cells, demonstrating an increased number of capillaries were present around the islet graft in the presence of MSCs compared to islet grafts without MSCs (191). VEGF and von Willebrand factor positive cells were localized after one week around the graft (192). A complete vascularization has been documented after subcutaneous transplantation in immunodeficient STZ-induced diabetic SCID mice after 84 days (186). MSCs migrated toward islet grafts to promote vascularization. After islet transplantation under the kidney capsule, MSCs infused by intraperitoneal injection migrated to the graft site and differentiated into vascular cells, producing VEGF (187).
32
4.2.3 Mechanism of islet-MSC cell-to-cell contact
It is known that cell-to-cell interactions within the islet cells, involving neural (N)- and epithelial (E)-cadherin, protect cells from apoptosis (199) and are also important to warrant sufficient insulin secretion after glucose stimulation (200). However, the mechanism of cell interaction between islets and MSCs is not clear.
Cellular adhesions and junctions are essential in the maintenance of the islet. Rho-associated kinase (ROCK) is responsible for cell polarity, morphology and motility (201) and its inhibition facilitate cellular stability, promoting adhesion through cadherins (202). In effect, the inhibition of ROCK in a co-culture of islets and MSCs compacted the structure of the heterogeneous co-culture (203). Moreover, islets cultured with MSC-conditioned medium expressed higher levels of pAKT and pERK and their neutralization totally inhibited islet proliferation capacities that were increased in the presence of MSCs (204). Elastin microfibril interface 1 (EMILIN-1), integrin-linked protein kinase (ILK) and hepatoma-derived growth factor (HDGF) positively correlate with the regenerative effects of MSCs (205). Furthermore, annexin A1 is involved in cell-to-cell contact between islets and MSCs (206). The addition of annexin A1 to islet culture in vitro mimics the effect of MSCs and increases insulin secretion;
moreover, silencing of annexin A1 by si-RNA or genetic mouse knock-out models decreased insulin secretion capacities (206).
These studies explain in part the interactions occurring between MSCs and islets; however the molecules involved in the cell-to-cell contact or the secreted molecules that improve islet viability have not been fully identified.
33
4.3 Mesenchymal stromal cells and liver cells
Hepatocytes are the principal components of the liver parenchyma and they are arranged in linear cords to form a liver lobule, which is defined at a histological scale as a small division of the liver. These lobules are surrounded by the portal triad, composed of the bile duct, and the interlobular vein and artery where blood enters and flows into sinusoids of the lobules.
Hepatocytes are responsible for the secretion of albumin, urea and drug metabolization, thus allowing medicaments to be eliminated (207).
4.3.1 MSCs sustain hepatocytes
MSCs furnish a stroma structure for hepatocytes and help to restore the lobes’ morphology present in the liver (208). Hence, MSCs co-cultured with isolated hepatocytes promoted cell stability (209). MSCs and hepatocytes cultured in a three-dimensional culture fashion showed increased hepatocyte stability, and thus, functionality as well (210). Hepatocyte viability was also improved in a co-culture model with bio-artificial materials, such as poly(lactic-co- glycolic) acid (PLGA) scaffold or engineered liver scaffolds. It has been suggested that these structures sustain hepatocytes and foster cellular adhesion. The presence of MSCs further improves hepatocyte function and survival, probably facilitating the establishment of cell-to- cell contact and cell-to-matrix interactions (211-213). In fact, MSCs most probably condition the microenvironment of hepatocytes through the secretion of collagen, laminin and fibronectin, dermatan and chondroitin sulfate-proteoglycans, which are the principal components of the extra cellular matrix (208). Moreover, the presence of MSCs improves hepatocyte polarization; in fact, MSCs increased hepatocyte functions during the first week of co-culture and restored their polarity as shown by transmission electron microscopy (214).
34
4.3.1 MSCs improve hepatocyte viability and functionality
Several studies suggest that MSCs promote hepatocyte viability and proliferation as demonstrated by co-culture of human adipose-derived MSCs with primary human hepatocytes (209). As shown by others, the expression of hepatotrophic and anti-apoptotic genes might be involved in the beneficial effects of MSCs (215). Indeed, in the co-culture models with MSCs, hepatic cells in the G2/S phase of the cell cycle increased in number, demonstrating that hepatocytes are driven to proliferate (16).
Furthermore, MSCs improve hepatocyte functionality (210, 216). Indeed, previously cryopreserved hepatocytes co-cultured with MSCs secreted increased levels of albumin and urea (216). A similar effect was observed after co-encapsulation in alginate-poly-L-lysine polymer or in spheroids, where hepatocytes presented a significantly improved secretion of albumin and urea (16, 210). Contrarily to this, even though other studies reported that albumin and urea secretion were not affected by the presence of MSCs; however, hepatocytes co- cultured with MSCs presented an improved morphology, phenotypic stability, as shown by the expression of CK18 and HNF4α, and increased gene expression of cytochrome P450 (209).
The mechanism involved in these effects remains obscure; but, there is a study showing that he production of IL-6 in MSC-hepatocyte co-cultures is significantly increased in contrast to hepatocytes that are cultured alone. Interestingly, IL-6 neutralization in vitro prevented the increased levels of albumin and urea synthesis, present in the hepatocyte and MSC co-culture, demonstrating the involvement of IL-6 in the observed functional effects. Further, no differences in TGF-α or TNF-α levels were detected in supernatants of hepatocytes cultured alone or those cultured with MSCs (217). To understand the mechanism leading to the increased function of hepatocytes co-cultured with MSCs, Huang et al cultured hepatocytes
35
with MSCs and induced cell damage through the addition of acetaminophen in the culture.
Effectively, acetaminophen induces cell death in hepatocytes via mitochondrial damage (218).
The presence of MSCs improved cell viability, which correlated with decreased levels of lactate dehydrogenase, an indicator of reduced cytotoxicity. Furthermore, the presence of phosphorylated JNK and ERK was assessed, and they were overexpressed after injury induction. This overexpression of JNK and ERK was suppressed in hepatocytes co-cultured with MSC. This suggests that hepatocyte damage caused by acetaminophen activates JNK and ERK pathways, and that the presence of MSCs can inhibit the activation of this signaling (219).
Hepatocytes co-transplantated with MSCs or treated with MSC-conditioned medium before transplantation, exert benefic effect on liver injury in vivo. In fact, mouse primary hepatocytes culture, pretreated with MSC-conditioned medium and then transplanted into CCL4-treated mice exhibited an attenuation of early apoptosis through the activation of fibroblast-like protein-1 (220). Also the co-transplantation of MSCs together with hepatocytes improved liver functions and survival rate in rats (16), as shown by decreased levels of the hepatic enzymes, ALT and AST, in mice (211). The beneficial effects demonstrated in vivo are suggested to be caused by paracrine molecules and by cell-to-cell contacts; however, the clear mechanisms involved in MSC and hepatocyte interactions leading to the beneficial effects are not well known.
36
4.4 Cell encapsulation
Cellular transplantation could represent an attractive and easier therapeutic solution compared to full organ transplantation. Clinical islet cell transplantation is a therapeutic option in T1D, with difficult management and hypoglycemia unawareness. In acute liver failure, cell transplantation could be a solution when full liver transplantation cannot be offered. The use of animals as cell donors represents a possible new source for tissues to overcome the issue of limited availability. Cell encapsulation has the advantage to protect transplanted cells from the immune system, while also permitting oxygen and nutrients to reach these encapsulated cells by diffusing in between the polymer. Successful encapsulation would obviate the problem of systemic immunosuppression in cell transplant recipients. Permeability of the polymer is various but normally does not exceed pore sizes that allow the passage of molecules of >/=
100 kDa, preventing antibodies or immune cells to penetrate into the capsule (Figure 3), however allowing oxygen, hormons and nutrients to pass through (221).
Figure 3. Bead properties.
37
Several polymers, such as agarose, chitosan and collagen, have been tested during recent decades and the polymer that remain widely used is alginate (222). Alginate is a hydrogel, being composed of a high quantity of water, and this renders the polymer biocompatible.
Nevertheless, alginate is not ideal since microcapsules are mechanically unstable, leading to desegregation of capsules and to the release of immunogenic material over time (223). For these reasons, studies are ongoing to develop more suitable polymers or linking alginate with other polymers, in particular with poly(ethylene glycol) (PEG) (224, 225). These more advanced polymers are more stable and suitable for cell encapsulation (226).
4.4.1 Islet encapsulation and transplantation outcome
Islet cell transplantation is used worldwide to treat unstable cases of T1D and the number of patients waiting for transplantation increases constantly. Life-long harmful immunosuppressive treatments are necessary. With the aim to avoid immunosuppression treatments and to use new unlimited cell sources, encapsulation studies have been performed with islets (227).
In a pig-to-primate xenotransplantation model encapsulated islets were incorporated into a monolayer device and transplanted subcutaneously. Control animals transplanted with free islets immediately rejected the graft; however, diabetic primates transplanted with encapsulated islets achieved and maintained normoglycemia up to six months without immunosuppressive treatments, until breakage of the microcapsules (228). Similarly, in another study, islets were encapsulated using the TheraCyte macro capsules and implanted subcutaneously in non-human primates. Diabetes in the recipient was induced by partial pancreatectomy hyperglycemia was reduced after transplantation of macro-encapsulated islets. After retrieval at 3, 6 and 12 months, the device presented viable islets, onset of
38 vascularization and minimal fibrosis (229). Clinical allo-transplantation using encapsulated islets has been performed in T1D patients. Encapsulated islets were injected in the peritoneal cavity; under such conditions patients continued to require exogenous insulin supply (230- 232). However, no infections or immune reactions were reported. Occasionally capsules were attached to the peritoneum, or formed fibrotic lumps (230, 231). In 2007, one patient underwent an intraperitoneal xenotransplantation of encapsulated porcine islets. Exogenous insulin supply was reduced by 30% during the first 12 weeks and returned to pre-transplant levels 49 weeks after transplantation (233).
Furthermore, several studies using various types of polymers demonstrated that co- encapsulation of rodent islets, together with MSCs, improved insulin secretion and increased graft survival in rodents after transplantation (234-237). Until today, only few studies investigated the effect of human MSCs on human islets, and their co-encapsulation and transplantation has, to our knowledge, not been described so far.
4.4.2 Hepatocyte encapsulation and transplantation outcome
Several types of microcapsules have been tested for hepatocyte encapsulation (238), using biomaterials such as alginate, polyurethane, chitosan and polycaprolactone. Hepatocytes adhered to all polymers, suggesting that the tested polymers conferred structural support to hepatocytes (239, 240).
Encapsulation of primary hepatocytes with alginate, PEG, chitosan, silk sericin-alginate- chitosan and three-layered alginate-chitosan-alginate polymers maintained typical morphology, albumin and urea secretion, and maintained their metabolic capacities for up to four months (241-244). Nevertheless, encapsulation with PEG resulted in major cell death one
39
day after encapsulation (245), demonstrating the technical challenge of primary hepatocytes encapsulation.
Baboons, transplanted with encapsulated pig hepatocytes, recovered from acute liver injury induced by 75% hepatectomy and warm ischemia. The intraperitoneal transplantation allowed 75% of baboons to recover from liver injury, while 25% developed liver failure after 21 days.
Control animals developed liver failure after six to ten days (246).
Co-encapsulation of rat hepatocytes together with MSCs showed enhanced albumin secretion (247, 248), improved ammonia metabolism and urea synthesis (249), decreased bilirubin levels (248) and improved graft survival and function for up to four months in allo- transplantation models (250, 251).
Previously, in our laboratory, an efficient technique of isolation and encapsulation of primary pig (252) and human (253) hepatocytes had been established. Until today, no standardized protocols for clinical applications are available. Further experimental studies transplanting encapsulated hepatocytes needs to be performed, to evaluate the potential therapeutic usefulness for the treatment of acute liver failure.
40
5. RESULTS
5.1 Aims
The aim of this thesis was to assess whether bone-marrow derived human MSCs have a supportive effect on the function of human islets of Langerhans and porcine hepatocytes.
These studies aimed to contribute to the elucidation of molecular mechanisms involved in the beneficial effects observed by using culture-expanded MSCs. Furthermore, one global objective was to analyze the usefulness of MSCs in survival and function of encapsulated islets and hepatocytes intended to cell therapies for the treatment of T1D and acute liver failure.
Specific Aim I:
The first aim was to analyze whether MSCs and human islet co-culture improves human β cell function, i.e. insulin secretion, using in vitro approaches. Further we aimed to analyze the molecular mechanisms such as cell-to-cell contact involved in the beneficial effects observed.
Additionally, we analyzed the effects of MSCs co-encapsulated with human islets on graft function and survival after transplantation in diabetic mice.
Specific Aim II:
The second aim was, firstly, to optimize a protocol to obtain high yield preparations of viable porcine hepatocytes and, secondly, to encapsulate hepatocytes in newly developed polymers for transplantation purposes. We further analyzed the effect of MSCs on free and encapsulated porcine hepatocyte viability, albumin secretion and metabolism in order to evaluate the usefulness of co-encapsulation of hepatocytes and MSCs for future clinical applications.
41
42
5.2 MSCs and human islets of Langerhans
“Multipotent mesenchymal stromal cells enhance insulin secretion from human islets via N-Cadherin interaction and prolong function of transplanted encapsulated islets in mice”
E. Montanari et al. 2017 (under revision in Stem Cell Research and Therapy)
MSCs co-cultured with human islets of Langerhans improved insulin secretion by cell-to-cell contact in vitro. The screening of adhesion molecules by qPCR in islets and in MSCs, revealed the expression of intercellular adhesion molecule (ICAM)-1 and N-cadherin in both cell types. Blocking of N-cadherin inhibited the increased insulin secretion in MSC-islets co- culture conditions, without affecting insulin secretion by islets cultured alone, demonstrating the involvement of N-cadherin interactions in the observed effect. Histology studies evidenced the contact of MSCs with islets, thus providing a cellular basement that surrounds islets and serves as a stromal structure. Co-encapsulation and co-transplantation of MSCs in vivo along with islets in diabetic mice significantly increased graft functionality and prolonged graft survival.
Personal contribution:
In this work, I designed and performed experiments, analyzed the data and wrote the manuscript.
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
